总结：我正在尝试使用 SQL PIVOT 将性别作为 AgeBand 的子类别（请注意，年龄带每个跨越两列）。我已经让 AgeBand 本身工作了（见下文），但不确定如何在 SQL PIVOT (SQL Server) 中添加第二个级别（性别）。
| 65-74 | 65-74 | 75-84 | 75-84 |
|---|---|---|---|
| 女 | 男 | 女 | 男 |
| 123 | 234 | 56 | 654 |
| 123 | 5345 | 456 | 76 |
我有一个 SQL 查询，它使用 PIVOT 从如下所示的表中获取数据：
| 县 | 年龄范围 | 性别 | 部分 | 代码 | 描述 | 数 | 百分比 |
|---|---|---|---|---|---|---|---|
| 亚特兰蒂斯 | 65-74 | F | 13.一般 | F34 | 苹果 | 279 | 0.089 |
| 亚特兰蒂斯 | 65-74 | M | 13.一般 | H84 | 橙色 | 166 | 0.053 |
| 亚特兰蒂斯 | 65-74 | F | 13.一般 | W234 | 梨子 | 8 | 0.003 |
| 亚特兰蒂斯 | 75-84 | M | 13.一般 | U7 | 李子 | 541 | 0.261 |
| 亚特兰蒂斯 | 65-74 | M | 13.一般 | V467 | 桃子 | 2 | 0.001 |
...等等...
看起来像这样（基于年龄范围的计数和百分比）：
| 县 | 部分 | 代码 | 描述 | 整体 | 整体 | 65-74 | 65-74 | 75-84 | 75-84 | 85+ | 85+ |
|---|---|---|---|---|---|---|---|---|---|---|---|
| - | - | - | - | 数 | % | 数 | % | 数 | % | 数 | % |
| 亚特兰蒂斯 | 13.一般 | F34 | 苹果 | 876 | 5 | 74 | 3 | 96 | 1 | 9 | 2 |
| 亚特兰蒂斯 | 13.一般 | H84 | 橙色 | 876 | 7 | 67 | 7 | 86 | 4 | 6 | 1 |
| 亚特兰蒂斯 | 13.一般 | W234 | 梨子 | 95 | 4 | 76 | 4 | 135 | 6 | 8 | 1 |
| 亚特兰蒂斯 | 13.一般 | U7 | 李子 | 986 | 7 | 23 | 7 | 5 | 3 | 8 | 2 |
| 亚特兰蒂斯 | 13.一般 | V467 | 桃子 | 1011 | 4 | 65 | 9 | 43 | 6 | 6 | 1 |
...
执行此操作的 SQL 效果很好：
SELECT County, Section, MeasureCode, MeasureDesc
, MAX(CASE WHEN theValue = 'Members' THEN CAST([Overall] as int) ELSE 0 END) as [Overall Members]
, MAX(CASE WHEN theValue = 'Percent' THEN CAST([Overall] as decimal(10,2)) ELSE 0 END) as [Overall Percent]
, MAX(CASE WHEN theValue = 'Members' THEN CAST([65-74] as int) ELSE 0 END) as [65-74 Members]
, MAX(CASE WHEN theValue = 'Percent' THEN CAST([65-74] as decimal(10,2)) ELSE 0 END) as [65-74 Percent]
, MAX(CASE WHEN theValue = 'Members' THEN CAST([75-84] as int) ELSE 0 END) as [75-84 Members]
, MAX(CASE WHEN theValue = 'Percent' THEN CAST([75-84] as decimal(10,2)) ELSE 0 END) as [75-84 Percent]
, MAX(CASE WHEN theValue = 'Members' THEN CAST([85+] as int) ELSE 0 END) as [85+ Members]
, MAX(CASE WHEN theValue = 'Percent' THEN CAST([85+] as decimal(10,2)) ELSE 0 END) as [85+ Percent]
FROM (
SELECT 'Members' as theValue
, County, Section, MeasureCode, MeasureDesc
, [Overall], [65-74], [75-84], [85+]
FROM #cond a
PIVOT(SUM(MemberCount)
FOR AgeBand IN ([Overall], [Under 65], [65-74], [75-84], [85+])
) b
UNION
SELECT 'Percent' as theValue
, County, Section, MeasureCode, MeasureDesc
, [Overall], [65-74], [75-84], [85+]
FROM #cond a
PIVOT(MAX(PctTotal)
FOR AgeBand IN ([Overall], [65-74], [75-84], [85+])
) c
) e
GROUP BY County, Section, MeasureCode, MeasureDesc
ORDER BY Section
, MAX(CASE WHEN Section = '13. Another Section' THEN CAST(MeasureDesc as int)
WHEN theValue = 'Percent' THEN CAST([Overall] as decimal(10,2))
WHEN theValue = 'Members' THEN CAST([Overall] as decimal(10,2))
ELSE 0
END) DESC
正如您所看到的，在原始数据集中，有一个性别列。我想将其添加为 AgeBands 的子类别，以便它看起来像这样（添加颜色和边框以使其更清晰）：
（请注意，每个年龄段部分现在都有女性和男性细分 - 抱歉，表格标记无法执行合并单元格，这将使这一点更加清晰）
| 县 | 部分 | 代码 | 描述 | 整体 | 整体 | 整体 | 整体 | 65-74 | 65-74 | 65-74 | 65-74 | 75-84 | 75-84 | 75-84 | 75-84 | 85+ | 85+ | 85+ | 85+ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| - | - | - | - | 女 | 女 | 男 | 男 | 女 | 女 | 男 | 男 | 女 | 女 | 男 | 男 | 女 | 女 | 男 | 男 |
| - | - | - | - | 数 | % | 数 | % | 数 | % | 数 | % | 数 | % | 数 | % | 数 | % | 数 | % |
| 亚特兰蒂斯 | 13.一般 | F34 | 苹果 | 876 | 5 | 876 | 5 | 74 | 3 | 74 | 3 | 96 | 1 | 96 | 1 | 9 | 2 | 9 | 2 |
| 亚特兰蒂斯 | 13.一般 | H84 | 橙色 | 876 | 7 | 876 | 7 | 67 | 7 | 67 | 7 | 86 | 4 | 86 | 4 | 6 | 1 | 6 | 1 |
| 亚特兰蒂斯 | 13.一般 | W234 | 梨子 | 95 | 4 | 95 | 4 | 76 | 4 | 76 | 4 | 135 | 6 | 135 | 6 | 8 | 1 | 8 | 1 |
| 亚特兰蒂斯 | 13.一般 | U7 | 李子 | 986 | 7 | 986 | 7 | 23 | 7 | 23 | 7 | 5 | 3 | 5 | 3 | 8 | 2 | 8 | 2 |
| 亚特兰蒂斯 | 13.一般 | V467 | 桃子 | 1011 | 4 | 1011 | 4 | 65 | 9 | 65 | 9 | 43 | 6 | 43 | 6 | 6 | 1 | 6 | 1 |
但是，我完全不知道如何在 SQL 中实现这一点。有人能指出我正确的方向吗？
我想 SQL 结果不能有 2 个列标题。但是您可以在代码中添加几行，如下所示？
SELECT County, Section, MeasureCode, MeasureDesc
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'F') THEN CAST([Overall] as int) ELSE 0 END) as [F_Overall Members]
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'M') THEN CAST([Overall] as int) ELSE 0 END) as [M_Overall Members]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'F') THEN CAST([Overall] as decimal(10,2)) ELSE 0 END) as [F_Overall Percent]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'M') THEN CAST([Overall] as decimal(10,2)) ELSE 0 END) as [M_Overall Percent]
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'F') THEN CAST([65-74] as int) ELSE 0 END) as [F_65-74 Members]
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'M') THEN CAST([65-74] as int) ELSE 0 END) as [M_65-74 Members]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'F') THEN CAST([65-74] as decimal(10,2)) ELSE 0 END) as [F_65-74 Percent]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'M') THEN CAST([65-74] as decimal(10,2)) ELSE 0 END) as [M_65-74 Percent]
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'F') THEN CAST([75-84] as int) ELSE 0 END) as [F_75-84 Members]
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'M') THEN CAST([75-84] as int) ELSE 0 END) as [M_75-84 Members]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'F') THEN CAST([75-84] as decimal(10,2)) ELSE 0 END) as [F_75-84 Percent]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'M') THEN CAST([75-84] as decimal(10,2)) ELSE 0 END) as [M_75-84 Percent]
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'F') THEN CAST([85+] as int) ELSE 0 END) as [F_85+ Members]
, MAX(CASE WHEN (theValue = 'Members' And MemberGender = 'M') THEN CAST([85+] as int) ELSE 0 END) as [M_85+ Members]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'F') THEN CAST([85+] as decimal(10,2)) ELSE 0 END) as [F_85+ Percent]
, MAX(CASE WHEN (theValue = 'Percent' And MemberGender = 'M') THEN CAST([85+] as decimal(10,2)) ELSE 0 END) as [M_85+ Percent]
FROM (
SELECT 'Members' as theValue
, County, Section, MeasureCode, MeasureDesc, MemberGender
, [Overall], [65-74], [75-84], [85+]
FROM #cond a
PIVOT(SUM(MemberCount)
FOR AgeBand IN ([Overall], [Under 65], [65-74], [75-84], [85+])
) b
UNION
SELECT 'Percent' as theValue
, County, Section, MeasureCode, MeasureDesc, MemberGender
, [Overall], [65-74], [75-84], [85+]
FROM #cond a
PIVOT(MAX(PctTotal)
FOR AgeBand IN ([Overall], [65-74], [75-84], [85+])
) c
) e
GROUP BY County, Section, MeasureCode, MeasureDesc, MemberGender
ORDER BY Section
, MAX(CASE WHEN Section = '13. Another Section' THEN CAST(MeasureDesc as int)
WHEN theValue = 'Percent' THEN CAST([Overall] as decimal(10,2))
WHEN theValue = 'Members' THEN CAST([Overall] as decimal(10,2))
ELSE 0
END) DESC